DECEMBER,
NOTES
1963
yellow liquid. Dried 12 hr. over phosphorlc anhydride, the liquid weighed 2.3661 g. Ita infrared epectrum waa identical to that of 1,3-propanesultam.
Acknowledgment.-We are indebted to Mr. P. Waszeciak, who prepared a sample of chromatographi-
3541
cally pure propanesultam by fractional melting. We also with thanks the assistance of Dr. Sidney Siggia, Director of the Olin Central Analytical Laboratories, and his staff for the microanalyses, infrared spectra, intrinsic viscosities, and V.P.C. results reported.
Notes The Condensation of Acetone with D i m e t h y l f u l v e n e l WILLIAM B. SMITHAND CARLOS GONZALEZ Department of Chemistry, Texas Christian University, Fort Worth, Texas Received April 1, 1963
While it has been recognized for many years that cyclopentadiene will condense with more than one mole of acetone under basic conditions, it is noteworthy of the complexity of the chemistry of these compounds that no adequate proof of structure of the condensation product@) has ever been given.2 Thiele and Balhorna reported that cyclopentadiene reacted with three equivalents of acetone and that dimethylfulvene reacted with two equivalents of acetone to give I, a structure assigned solely on the basis of elemental analysis.
Subsequently, Ziegler and Crossman4 modified the structure to either I1 or I11 on the basis of the related reaction of dimethylfulvene with anisaldehyde. Their argument rested on the observation that the condensation product of anisaldehyde with dimethylfulvene did not involve reaction with the methyl groups a t the 6position. Only one mole equivalent of anisaldehyde entered into the reaction. The preparation of dimethylfulvene by the reaction of cyclopentadiene and acetone in the presence of alcoholic potassium hydroxide5 always leads to the formation of a considerable amount of higher boiling material. Careful fractional distillation of the material a t reduced pressures produced one main product (an orange oil) plus a large amount of tar. The same product was isolated when dimethylfulvene reacted (1) Presented at the Southwest Regional Meeting of the American Chemical Society, Dallas. Tex., December 7, 1962. (2) J. H. Day, Chem. Rev.. 68, 167 (1953). (3) J. Thiele and H. Balhorn. Ann., 848, 1 (1906). (4) K.Ziegler and F. Crossman, ibid., 611, 89 (1934). (5) G. Crane, C. E. Board, and A. L. Henne, J . 070. Chem., 87, 1237 (1945).
with acetone in alcoholic potassium hydroxide. As will be seen subsequently, this product resulted from the addition of one mole of acetone to dimethylfulvene. Numerous repetitions of the preparation of dimethylfulvene and of the subsequent reaction of this with acetone failed to yield any evidence of a product corresponding to the condensation of two molecules of acetone per one molecule of dimethylfulvene. The chemical analysis and hydrogenation equivalent of the condensation product from dimethylfulvene and acetone established the formula as CllHlaO with three double bonds in the molecule. The infrared spectrum indicated the presence of a hydroxyl group. The hydrogenated product was a colorless oil, the analysis of which agrees with the formula C11H220. This material also had a strong hydroxyl band in the infrared. The position of attachment was elucidated by subjecting the hydrogenated product to dehydration and ozonolysis yielding 3-isopropylcyclopentanone. While these data locate the position of the acetone moiety a t C-2 on the fulvene ring, the possible location of the three double bonds requires further information. The fact that the ultraviolet absorption spectra of the condensation product and of dimethylfulvene show maxima a t 267 mp ( e 1.8 X lo4 and 2.2 X lo4, respectively) suggests similar conjugated structures for both compounds. This would give structure IV as the correct formula for the condensation product. FHs
Confirmation of the proposed structure was obtained from the proton magnetic resonance spectrum of the condensation product. The following absorptions and their respective integrated proton values were observed in a 10% carbon tetrachloride solution: (1) the ring protons consisting of doublet a t 7 3.61 (two protons) and a triplet at T 3.86 (one proton); (2) a broad weak peak at 7 7.15 which was variable in position with dilution was attributed to the hydroxyl hydrogen; (3) a sharp band a t 7 7.91 (6 protons); and (4) a similar band a t T 8.61 (6 protons). A detailed examination of the ring-H resonances revealed a typical AB2 spectrum where UA < UB; v06 = -13.5 C.P.S. The observed spectrum was analyzed by means
NOTES
3542
of Corio's tables.6 A good fit for the AB2 case for both ~ line position and intensity was found for J A B / v ~equal to 0.15. On this basis, the value of J A B = 2.0 C.P.S. These data are all consistent with the structure of the condensation product as 2-(l-hydroxy-l-methylethyl)6,6-dimethylfulvene (IV).' Experimental Reaction of Acetone with Cyc1opentadiene.-The reaction of 181 g. of freshly prepared cyclopentadiene, 160 g. of acetone, and 69 ml. of 20'30 potassium hydroxide in ethanol was carried out as directed by Crane, Boord, and Henne.6 After standing overnight in an icebox, the mixture was poured into a large volume of water. The total mixture was then extracted with several portions of ether. The combined extract was dried over magnesium sulfate, and, after filtration, the ether was removed by distillation through a 34-cm. Vigreux column. The residue was then distilled through the same column a t 8 mm. yielding the following fractions: 124 g. of dimethylfulvene, b.p. 4&43"; and 9.7 g. of orange oil with 12% 1.5425, b.p. 118-120". The remaining material was a nondistillable tar. Reaction of Dimethylfulvene with Acetone .-Dimethylfulvene (123 g.), 102 g. of acetone, and 100 ml. of ll%>otassium hydroxide in ethanol reacted with stirring a t 0' overnight. The reaction mixture was worked up as described in the preceding reaction, and the product distilled (8 mm.) as follows: b.p. 110-124" (mostly 118-124'), 4.7 g. of orange oil with n26D 1.5438. The infrared spectra of this material was identical with that of the material reported in the previous reaction. Characterization and Degradation of the Orange Oil.-The orange oil, b.p. 118-120' (8 mm.), n% 1.5425, showed a strong hydroxyl band in the infrared a t 2.98 p . The hydrogenation equivalent was determined using Adam's -platinum catalyst in ethanol: calcd. for CllHlcO with three double bonds, 54.8; found, 56.6. The compound in ethanol showed a maximum at 267 mp ( e 2.2 X 104)). The n.m.r. spectrum was determined in carbon tetrachloride solution (tetramethylsilane as an internal reference) with a Varian A-60 instrument, and the observed absorptions are given in the text. Anal.8 Calcd. for CllHlaO: C, 80.44; H , 9.82. Found: C, 80.59; H , 9.80. A sample of this material was readily hydrogenated in ethanol using a 5% palladium-on-charcoal catalyst. Work-up of the product in the usual fashion and distillation through a semimicro column gave a colorless oil, b.p. 83-86' (6 mm.), 12% 1.4670, with a strong band in the infrared a t 2.92 p . Anal. Calcd. for CllHtzO: C, 77.58; H , 13.02. Found: C, 77.74; H, 13.13. The Saturated alcohol (6 g. of alcohol mixed with 0.6 g. of piperadine) was dehydrated by slow passage in a stream of nitrogen through a vertical 18-cm. column of 8-mesh alumina heated to ca. 355". The product condensed as a light yellow oil which was taken up in ether, washed with dilute hydrochloric acid, and dried over magnesium sulfate. After filtration, the ether was removed by distillation through a short Vigreux column. The residue was distilled through a semimicro Vigreux column to give a mixture of olefins (2.8 g.), b.p. 146-150'. The strong infrared band a t 11.3 p indicated that the major component contained a terminal double bond and, from subsequent data, was presumably 3-isopropenylisopropylcyclopentane. This component was isomerized to 3-isopropylideneisopropylcyclopentane by treating the olefin mixture on a steam bath for 2 hr. withO.l g. of p-toluenesulfonic acid. I t was noted a t this time that the 11.3-/.Iband had disappeared from the spectrum. The crude olefin mixture was taken up in 25 ml. of methylene dichloride, washed with diluted sodium bicarbonate, and ozonized in a stream of ozonized air until saturated. Excess ozone was removed in a stream of nitrogen. The ozonide(s) was decomposed by stirring the solution with 3 ml. of 30% hydrogen peroxide in 22 ml. of 10% potassium hydroxide. Water (2.5 ml.) was added, and the organic layer was removed. The aqueous layer was (6) P. L. Corio, Chem. Reu.. 6 0 , 3 6 3 (1960). (7) Recently, D. M. Fenton, and M..J. Hurwitz, J . O7g. Chem,, 38, 1646 (1963), established t h a t , when dimethylfulvens and acetone are refluxed in methanolic potassium hydroxide, one obtains 3-isopropenyl-6.6-dimethylfulvene. a result in agreement with o u r findings. ( 8 ) Microanalysis b y Galbraith Laboratories, Knoxville, Tenn.
VOL.28
further extracted with methylene dichloride. The methylene dichloride extracts were combined and dried, and the solvent was removed by distillation. An authentic sample of 3-isopropylcyclopentanone was prepared by the hydrogenation of 3-isopropenylcyclopentanone8 using Adam's platinum in ethyl acetate: b.p. 94-97" (34 mm.), n 2 b 1.4461; lit.lo b.p. 183' (761 mm.), 12% 1.4443. The infrared spectrum of the compound indicated the absence of the unsaturated starting material. Analysis by V.P.C. indicated purity. The crude ozonization product was fractionated by V.P.C. using a 12-ft. silicone column. In this fashion, it was shown that the reaction mixture consisted of a major band corresponding in retention time to authentic 3-isopropylcyclopentanone plus several smaller bands. The band corresponding in retention time to authentic 3-isopropylcyclopentanone was trapped, R ~ ~ 1.4443. The infrared spectrum of this material was identical with that of the autheutic sample; strong bands were observed a t 1750, 1475, 1412, 1390, 1375, 1160, and 820 em.-'. For comparison purposes an authentic sample of 2-isopropylcyclopentanone was prepared by the method of Cornubert and Borrel." The retention time and infrared spectrum were decidedly different from the 3-isopropyl isomer (strong bands a t 1750, 1475, 1412, 1390, 1375, 1275, 1195, 1160, 935, and 830 cm.-l).
Acknowledgment.-This work was supported by a grant from the Petroleum Research Fund administered by the American Chemical Society. Grateful acknowledgment is extended to the donors of this fund. We also would like to express our appreciation to the Robert A. Welch Foundation whose generosity made available the Varian A-60. (9) C. Djerassi and B. Tursch, J . Org. Chem.. 88, 4609 (1961). (10) G.Burger and A. K. Macbeth, J . Chem. Soc.. 145 (1946). (11) R. Cornubert and C . Borrel, B d l . 8oc. chim. France, 958 (1930).
Dipicolinoyl- and Di(6-acetylpicolinoyl)furoxansl M. S. CHANG Research and Development Department, U . S . Naval Propellant Plant, Indian Head, Maryland Received April 30, 1963
The reaction of acetophenone and nitric acid has been known since 1887.2 The formation of dibenzoylfuroxan as the main product by dimerization of benzoylnitrile N-oxide has been proposed, and recently a minor product in this reaction has been assigned the structure of the dibenzoate ester of bis(benzoy1forma1doximino)furoxan.4 In similar work with thiophene derivatives, bis(3-thianaphthenoyl)furoxan5 and di(2-thenoyl)furoxan6 have been prepared. The present work has been directed towards a study of the reaction of nitric acid with acetylpyridine isomers and 2,6diacetylpyridine. As weak bases, acetylpyridine isomers form salts with mineral acids. Experimental results indicate that the formation of acetylpyridine nitrates occurs along with nitrosation and oxidation. From either 3- or 4-acetylpyridine and nitric acid, both the formation of the corresponding nitrate salt and oxidation to the cor(1) This work was supported b y the Foundational Research Program of the Bureau of Naval Weapons. (2) A. F. Holleman, Ber., 30, 3359 (1887). ( 3 ) H.R . Snyder and N. E. Boyer. J . A m . Chem. SOC.,77, 4233 (1955). (4) J. H.I3oyer and M . 8 . Chang, i b i d . , 83, 2220 (1960). (5) D. A. Shirley, B. H. Gross, and M. J. Dansig, J. O T ~Chem., . 38, 1024 (1958). (6) M.S.Chang and A. J. Matuazko, zbzd., 16,6239(1961).
D
